The Battle Continues: Facing the obstacle of tumor resistance to immunotherapy treatment
The immune system and cancer cells have interactions that are continuously evolving from the initial establishment of cancer to the development of metastatic disease. As research has shown, the complex cancer ecosystem enables numerous, devious ways for tumors to evade treatment. However, transformational strategies to prevent or treat the mechanisms of resistance to immunotherapy are being developed to further improve patient outcomes.
Hot or Cold?
One effort within the fight against tumor resistance is to identify biomarkers that can be used to delineate immunologically “cold” tumors from “hot” tumors. For example, from both mouse model work as well as human studies, it has been suggested that inactivation of STK11 is associated with a cold tumor microenvironment as evidenced by the reduced infiltration of cytotoxic lymphocytes. Building on this finding, investigators at MD Anderson hypothesized that STK11 genomic alterations may therefore predict lack of clinical benefit to anti-PD-1/PD-L1 treatment (Skoulidis et al 2018). They analyzed several cohorts of KRAS-mutant lung adenocarcinoma (LUAC) patients and found that mutant STK11 was a driver of resistance to checkpoint inhibitor treatment. The authors concluded that while there are still many determinants of response to immunotherapy treatment to explore, STK11 status might also be considered along with more established markers like TMB and PD-L1 expression in evaluating LUAC patients.
Further, the figure below illustrates why it is important to ensure the assays selected for biomarker discovery are accurate and comprehensive. As shown, sequencing coverage gaps or inconsistencies (regions with no or little blue coverage) exist in STK11 gene using a standard, off-the-shelf exome kit. Several of the mutations used to predict resistance by Skoulidis et al. were found in exon 8 (green region with enhanced coverage highlighted by red rectangle) and would be undetected by a conventional whole exome sequencing approach. Having a high content, broad approach for investigating markers of response is essential, but equally important is one that is accurate.
ACE Technology Fills Sequencing Coverage Gaps in Genomic Regions Associated with Biomarker Discovery
Modeling Tumor Immune Evasion to Investigate Resistance
Recent studies by researchers at the Dana-Farber Cancer Institute investigating transcriptomic signatures provide additional insights into tumor resistance against checkpoint blockade (Jiang et al. 2018). They developed an algorithm, TIDE, which integrates expression signatures of T cell dysfunction and exclusion to simulate tumor immune evasion. From analysis of pre-treatment samples with clinical outcome data available and preclinical model studies, they identified SERPINB9 as a potential new candidate for resistance to checkpoint inhibitors. SERPINB9, a protease inhibitor, is thought to promote resistance to T cell-mediated killing of tumor cells when overexpressed in cancer. Functional studies performed by knocking out SERPINB9 in cancer cell lines using CRISPR-Cas9 and in vitro cell cultures treated with IFN-y further corroborated the role of SERPINB9 in regulating resistance. Interestingly, the authors mention that there are no current small molecule inhibitors for SERPINB9, yet databases such as Pfizer’s OASIS indicates that this target is druggable. Cutting-edge studies like these can help the design of more innovative targets and drug combination therapies to improve patient treatment efficacy and hopefully diminish cancer progression.
What’s the best strategy for Immunotherapy combos to tackle tumor evasion?
At the Rational Combinations 360 meeting in Philadelphia in September, vital aspects of immunotherapy combinations were highlighted with many excellent presentations and scientific discussions in regard to translational science and emerging biomarkers. Lei Zheng, from Johns Hopkins University, presented on Overcoming the Resistance to Immune Checkpoint Inhibitors via Rational Combination of Immunotherapy and proposed his “idealized picture of combination immunotherapy.” Dr. Zheng suggested a 3-pronged approach consisting of 1) primers such as cytotoxics, cytokines, and vaccines; 2) checkpoints like PD-1 and CTLA-4, and 3) expanders that can overcome the immunosuppressive tumor microenvironment. Likewise, Daniel Chen from IGM Biosciences, touched on a similar theme of a multidimensional therapeutic strategy and equated this to putting “together an orchestra.”
An insightful panel discussion further asked, what specific action could we take to better understand mechanisms of resistance to treatment? Panelists mentioned the following as fundamental:
- Improvements in mouse models: These models need to be better representative and show the true principles.
- Availability of Pre- and Post- treatment samples: Make sure specimen are obtained and added into the archives and study design.
- Broad platforms are needed in translational research: Comprehensive assays are needed to explore all facets of tumor evasion.
These types of analyses and discoveries from refined study designs are possible with high content all-encompassing biomarker assays specialized for Immuno-Oncology.
Exploring mechanisms of tumor resistance requires comprehensive immuno-genomic profiling
In a recent interesting case study from the Personalis R&D team, we found highly suggestive resistance mechanisms to anti-PD-1 response. Our analysis highlights the importance of broad tumor immuno-genomic profiling in identifying resistance mechanisms in patients receiving immunotherapy. Through a collaborative study with Dr. Sekwon Jang at the INOVA Schar Cancer Institute in Virginia, we applied our ACE ImmunoID solution to sequence and profile tumors from 19 stage III and IV melanoma patients where response was evaluated using RECIST criteria. Of the 19 patients, there were 8 complete responders (CR), 2 partial responders (PR), and 9 progressive disease (PD) patients. The molecular data for each of the melanoma patient samples was analyzed together with the corresponding clinical response to anti-PD-1 therapy. Assessment of the neoantigen burden of these patients demonstrated that an overall, higher neoantigen load was associated with a better response to anti-PD1 therapy. However, two patient exceptions were analyzed further (green circles indicated in the figure below).
Higher Neoantigen Burden is Associated with a Better Response to Anti-PD-1 Therapy
In the patient with progressive disease, we uncovered that this individual has two likely very damaging class I HLA mutations; a somatic mutation in HLA-A0201 (stop gain mutation) and in HLA B1501 (a splice region variant). As HLA class I genes are major members of the antigen presentation machinery and necessary for presenting neo-epitopes in tumor cells to the immune system, these somatic HLA class I mutations likely play a role in the patients lack of response to anti PD-1 therapy. Secondly, after investigating the transcriptome results from the patient with partial response (yet high neoantigen load), this particular patient had very high IDO1 expression. It’s been shown that high IDO1 expression can lead to multiple immuno-suppressive actions, such as increased proliferation of immunosuppressive Tregs. For this particular patient, an Anti-IDO1 therapy, in combination with anti-PD1 treatment, could potentially improve the outcome in treatment.
These results demonstrate how an augmented Whole Exome and Whole Transcriptome approach can enable identification of potential mechanisms of tumor escape. In the past few years, our understanding of immuno-oncology has advanced quite substantially. While various tumor resistance obstacles can hinder the success of therapies, the development of more personalized solutions and innovations for drug combinations suggest that more progress is still to come.